Re: [PATCHv4 27/28] firmware: Add CONFIG_BUILTIN_FIRMWARE option

!MAILaRCHIVE_VOTE_RePLACE
Previous message: [thread] [date] [author]
Next message: [thread] [date] [author]
To: <linux-kernel@...>
Cc: <sam@...>, <alan@...>, <akpm@...>, <johannes@...>
Date: Wednesday, May 28, 2008 - 7:46 am

More feedback... Andrew wanted the help text to be a little more...
well, helpful. And Johannes wanted to add a CONFIG_BUILTIN_FIRMWARE_DIR
option, which lets you just point it at /lib/firmware or somewhere else
outside the kernel tree. And we fixed up a bunch of other details, like
the dependencies on directories, and the fact that the .S file doesn't
actually need to change when the binary blob does.

I even made the Makefile fit in 80 characters. It's going to make your
brain hurt either way -- but that's normal for kernel Makefiles.

commit 4311743f8ef92ab462c06a9dd2ff8e0d735b3dae
Author: David Woodhouse <dwmw2@infradead.org>
Date:   Fri May 23 13:58:12 2008 +0100

    firmware: Add CONFIG_BUILTIN_FIRMWARE option
    
    This allows arbitrary firmware files to be included in the static kernel
    where the firmware loader can find them without requiring userspace to
    be alive.
    
    (Updated and CONFIG_BUILTIN_FIRMWARE_DIR added with lots of help from
    Johannes Berg).
    
    Signed-off-by: David Woodhouse <dwmw2@infradead.org>
    Signed-off-by: Johannes Berg <johannes@sipsolutions.net>

diff --git a/Makefile b/Makefile
index 20b3235..ac2ab7e 100644
--- a/Makefile
+++ b/Makefile
@@ -450,7 +450,7 @@ scripts: scripts_basic include/config/auto.conf
 
 # Objects we will link into vmlinux / subdirs we need to visit
 init-y		:= init/
-drivers-y	:= drivers/ sound/
+drivers-y	:= drivers/ sound/ firmware/
 net-y		:= net/
 libs-y		:= lib/
 core-y		:= usr/
diff --git a/drivers/base/Kconfig b/drivers/base/Kconfig
index d7da109..f204ae7 100644
--- a/drivers/base/Kconfig
+++ b/drivers/base/Kconfig
@@ -34,6 +34,39 @@ config FW_LOADER
 	  require userspace firmware loading support, but a module built outside
 	  the kernel tree does.
 
+config BUILTIN_FIRMWARE
+	string "Firmware blobs to build into the kernel binary"
+	depends on FW_LOADER
+	help
+	  This option allows firmware to be built into the kernel, for the
+	  cases where the user either cannot or doesn't want to provide it from
+	  userspace at runtime (for example, when the firmware in question is
+	  required for accessing the boot device, and the user doesn't want to
+	  use an initrd).
+
+	  This option is a string, and takes the (space-separated) names of the
+	  firmware files -- the same names which appear in MODULE_FIRMWARE()
+	  and request_firmware() in the source. These files should exist under
+	  the directory specified by the BUILTIN_FIRMWARE_DIR option, which is
+	  by default the firmware/ subdirectory of the kernel source tree.
+
+	  So, for example, you might set CONFIG_BUILTIN_FIRMWARE="usb8388.bin",
+	  copy the usb8388.bin file into the firmware/ directory, and build the
+	  kernel. Then any request_firmware("usb8388.bin") will be
+	  satisfied internally without needing to call out to userspace.
+
+config BUILTIN_FIRMWARE_DIR
+	string "Firmware blobs root directory"
+	depends on BUILTIN_FIRMWARE != ""
+	default "firmware"
+	help
+	  This option controls the directory in which the kernel build system
+	  looks for the firmware files listed in the BUILTIN_FIRMWARE option.
+	  The default is the firmware/ directory in the kernel source tree,
+	  but by changing this option you can point it elsewhere, such as
+	  the /lib/firmware/ directory or another separate directory
+	  containing firmware files.
+
 config DEBUG_DRIVER
 	bool "Driver Core verbose debug messages"
 	depends on DEBUG_KERNEL
diff --git a/firmware/Makefile b/firmware/Makefile
new file mode 100644
index 0000000..c79bbe7
--- /dev/null
+++ b/firmware/Makefile
@@ -0,0 +1,86 @@
+#
+# kbuild file for firmware/
+#
+
+# Create $(fwdir) from $(CONFIG_BUILTIN_FIRMWARE_DIR) -- if it doesn't have a
+# leading /, it's relative to $(srctree).
+fwdir := $(subst ",,$(CONFIG_BUILTIN_FIRMWARE_DIR))
+fwabs := $(addprefix $(srctree)/,$(filter-out /%,$(fwdir)))$(filter /%,$(fwdir))
+
+fw-external-y := $(subst ",,$(CONFIG_BUILTIN_FIRMWARE))
+
+firmware-y    := $(fw-external-y) $(fw-shipped-y)
+firmware-dirs := $(sort $(patsubst %,$(objtree)/$(obj)/%/,$(dir $(firmware-y))))
+
+quiet_cmd_mkdir = MKDIR   $(subst $(srctree)/,,$@)
+      cmd_mkdir = mkdir -p $@
+
+quiet_cmd_ihex  = IHEX    $@
+      cmd_ihex  = objcopy -Iihex -Obinary $< $@
+
+ifdef CONFIG_64BIT
+ASM_WORD := .quad
+ASM_ALIGN := 3
+else
+ASM_WORD := .long
+ASM_ALIGN := 2
+endif
+
+quiet_cmd_fwbin = MK_FW   $@
+      cmd_fwbin = FWNAME="$(patsubst firmware/%.S,%,$@)";		     \
+		  FWSTR="$(subst /,_,$(subst .,_,$(patsubst 		     \
+				firmware/%.S,%,$@)))";			     \
+		  echo "/* Generated by firmware/Makefile */"		> $@;\
+		  echo "    .section .rodata"				>>$@;\
+		  echo "    .align $(ASM_ALIGN)"			>>$@;\
+		  echo "_fw_$${FWSTR}_bin:"				>>$@;\
+		  echo "    .incbin \"$(2)\""				>>$@;\
+		  echo "_fw_end:"					>>$@;\
+		  echo "   .section .rodata.str,\"aMS\",@progbits,1"	>>$@;\
+		  echo "    .align $(ASM_ALIGN)"			>>$@;\
+		  echo "_fw_$${FWSTR}_name:"				>>$@;\
+		  echo "    .string \"$$FWNAME\""			>>$@;\
+		  echo "    .section .builtin_fw,\"a\",@progbits"	>>$@;\
+		  echo "    .align $(ASM_ALIGN)"			>>$@;\
+		  echo "    $(ASM_WORD) _fw_$${FWSTR}_name" 		>>$@;\
+		  echo "    $(ASM_WORD) _fw_$${FWSTR}_bin" 		>>$@;\
+		  echo "    $(ASM_WORD) _fw_end - _fw_$${FWSTR}_bin"	>>$@;
+
+# One of these files will change, or come into existence, whenever
+# the configuration changes between 32-bit and 64-bit. The .S files
+# need to change when that happens.
+wordsize_deps := $(wildcard include/config/64bit.h include/config/32bit.h \
+		include/config/ppc32.h include/config/ppc64.h \
+		include/config/superh32.h include/config/superh64.h \
+		include/config/x86_32.h include/config/x86_64.h)
+
+# For the $$(dir %) trick, where we need % to be expanded first.
+.SECONDEXPANSION:
+
+$(patsubst %,$(obj)/%.S, $(fw-shipped-y)): %: $(wordsize_deps) \
+		| $(objtree)/$$(dir %)
+	$(call cmd,fwbin,$(patsubst %.S,%,$@))
+$(patsubst %,$(obj)/%.S, $(fw-external-y)): %: $(wordsize_deps) \
+		include/config/builtin/firmware/dir.h | $(objtree)/$$(dir %)
+	$(call cmd,fwbin,$(fwabs)/$(patsubst $(obj)/%.S,%,$@))
+
+# The .o files depend on the binaries directly; the .S files don't.
+$(patsubst %,$(obj)/%.o, $(fw-shipped-y)): %.o: %
+$(patsubst %,$(obj)/%.o, $(fw-external-y)): $(obj)/%.o: $(fwdir)/%
+
+$(obj)/%: $(obj)/%.ihex | $(objtree)/$(obj)/$$(dir %)
+	$(call cmd,ihex)
+
+$(firmware-dirs):
+	$(call cmd,mkdir)
+
+obj-y := $(patsubst %,%.o, $(firmware-y))
+
+# Remove .S files and binaries created from ihex
+# (during 'make clean' .config isn't included so they're all in $(fw-shipped-))
+targets := $(patsubst $(obj)/%,%, $(shell find $(obj) -name \*.S 2>/dev/null))
+targets += $(fw-shipped-)
+
+# Without this, built-in.o won't be created when it's empty, and the
+# final vmlinux link will fail.
+obj-n := dummy


-- 
dwmw2

--
Previous message: [thread] [date] [author]
Next message: [thread] [date] [author]

Messages in current thread:
[PATCH] keyspan_pda: use request_firmware(), David Woodhouse, (Fri May 30, 8:19 am)
[PATCH] keyspan: use request_firmware(), David Woodhouse, (Fri May 30, 7:45 am)
Re: [PATCH] keyspan: use request_firmware(), Hugh Blemings, (Fri May 30, 9:54 am)
[PATCH] ttusb-budget: use request_firmware(), David Woodhouse, (Thu May 29, 1:49 pm)
Re: [PATCH] ttusb-budget: use request_firmware(), Mauro Carvalho Chehab, (Fri Jun 13, 3:23 pm)
[PATCH] kaweth: use request_firmware(), David Woodhouse, (Thu May 29, 10:20 am)
Re: [PATCH] kaweth: use request_firmware(), Oliver Neukum, (Fri May 30, 9:55 am)
Re: [PATCH] kaweth: use request_firmware(), David Woodhouse, (Fri May 30, 11:16 am)
Re: [PATCH] kaweth: use request_firmware(), Marcel Holtmann, (Fri May 30, 11:26 am)
Re: [PATCH] kaweth: use request_firmware(), Alan Cox, (Fri May 30, 11:48 am)
Re: [PATCH] kaweth: use request_firmware(), Marcel Holtmann, (Sat May 31, 2:51 am)
Re: [PATCH] kaweth: use request_firmware(), Alan Cox, (Sat May 31, 8:35 am)
Re: [PATCH] kaweth: use request_firmware(), Oliver Neukum, (Fri May 30, 5:02 pm)
Re: [PATCH] kaweth: use request_firmware(), Alexandre Oliva, (Fri May 30, 8:00 pm)
[PATCH] smctr: use request_firmware(), David Woodhouse, (Thu May 29, 9:47 am)
Re: [PATCH] smctr: use request_firmware(), maximilian attems, (Thu May 29, 3:30 pm)
Re: [PATCH] smctr: use request_firmware(), David Woodhouse, (Thu May 29, 5:02 pm)
[PATCH N/28] ymfpci: treat firmware data as const, David Woodhouse, (Thu May 29, 8:21 am)
[PATCHv2 #ERROR!/28] maestro3: treat firmware data as const, David Woodhouse, (Thu May 29, 5:09 am)
[PATCHv2 11/28] cxacru: treat firmware data as const, David Woodhouse, (Sun May 25, 6:23 am)
[PATCHv2 14/28] rt2x00: treat firmware data as const, David Woodhouse, (Sun May 25, 6:23 am)
[PATCHv2 21/28] dvb frontends: treat firmware data as const, David Woodhouse, (Sun May 25, 6:23 am)
[PATCHv2 25/28] firmware: make fw-&gt;data const, David Woodhouse, (Sun May 25, 6:23 am)
[PATCHv2 18/28] cxgb3: treat firmware data as const, David Woodhouse, (Sun May 25, 6:23 am)
[PATCHv2 04/28] cx25840: treat firmware data as const, David Woodhouse, (Sun May 25, 6:23 am)
Re: [PATCHv2 04/28] cx25840: treat firmware data as const, Tyler Trafford, (Sun May 25, 12:16 pm)
Re: [PATCHv2 04/28] cx25840: treat firmware data as const, David Woodhouse, (Sun May 25, 10:56 am)
Re: [PATCHv2 04/28] cx25840: treat firmware data as const, Hans Verkuil, (Sun May 25, 11:04 am)
Re: [PATCHv2 04/28] cx25840: treat firmware data as const, David Woodhouse, (Sun May 25, 10:36 am)
[PATCHv2 01/28] libertas: treat firmware data as const, David Woodhouse, (Sun May 25, 6:23 am)
[PATCHv2 20/28] ttusb-dec: treat firmware data as const, David Woodhouse, (Sun May 25, 6:23 am)
[PATCHv2 09/28] vx: treat firmware data as const, David Woodhouse, (Sun May 25, 6:23 am)
[PATCHv2 19/28] bt8xx: treat firmware data as const, David Woodhouse, (Sun May 25, 6:23 am)
[PATCHv2 17/28] irda-usb: treat firmware data as const, David Woodhouse, (Sun May 25, 6:23 am)
[PATCHv2 15/28] p54: treat firmware data as const, David Woodhouse, (Sun May 25, 6:23 am)
[PATCHv2 03/28] cyclades: treat firmware data as const, David Woodhouse, (Sun May 25, 6:23 am)
[PATCHv2 27/28] firmware: Add CONFIG_BUILTIN_FIRMWARE option, David Woodhouse, (Sun May 25, 6:23 am)
[PATCHv3 27/28] firmware: Add CONFIG_BUILTIN_FIRMWARE option, David Woodhouse, (Sun May 25, 1:00 pm)
Re: [PATCHv4 27/28] firmware: Add CONFIG_BUILTIN_FIRMWARE op..., David Woodhouse, (Wed May 28, 7:46 am)
[PATCH 29/28] firmware: Add 'firmware_install' make target, David Woodhouse, (Thu May 29, 4:29 am)
[PATCHv2 22/28] cxusb: treat firmware data as const, David Woodhouse, (Sun May 25, 6:23 am)
Re: [PATCHv2 22/28] cxusb: treat firmware data as const, Michael Krufky, (Sun May 25, 10:14 am)
Re: [PATCHv2 22/28] cxusb: treat firmware data as const, David Woodhouse, (Sun May 25, 11:02 am)
Re: [PATCHv2 22/28] cxusb: treat firmware data as const, David Woodhouse, (Sun May 25, 10:27 am)
[PATCHv2 24/28] tuners: treat firmware data as const, David Woodhouse, (Sun May 25, 6:23 am)
[PATCHv2 23/28] gp8psk: treat firmware data as const, David Woodhouse, (Sun May 25, 6:23 am)
[PATCHv2 16/28] atmel: treat firmware data as const, David Woodhouse, (Sun May 25, 6:23 am)
[PATCHv2 12/28] aic94xx: treat firmware data as const, David Woodhouse, (Sun May 25, 6:23 am)
[PATCHv2 10/28] ueagle-atm: treat firmware data as const, David Woodhouse, (Sun May 25, 6:23 am)
[PATCHv2 13/28] zd1201: treat firmware data as const, David Woodhouse, (Sun May 25, 6:23 am)
[PATCHv2 08/28] pcxhr: treat firmware data as const, David Woodhouse, (Sun May 25, 6:23 am)
[PATCHv2 05/28] myri10ge: treat firmware data as const, David Woodhouse, (Sun May 25, 6:23 am)
[PATCHv2 07/28] riptide: treat firmware data as const, David Woodhouse, (Sun May 25, 6:23 am)
[PATCHv2 02/28] bluetooth: treat firmware data as const, David Woodhouse, (Sun May 25, 6:23 am)
[PATCHv2 06/28] vx222: treat firmware data as const, David Woodhouse, (Sun May 25, 6:23 am)